INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

EXTRACELLULAR NUCLEOTIDES can bind to P2-type purinoceptors, which constitute a large family of receptors that are ion channels (P2X receptors) or that couple to trimeric G proteins (P2Y receptors) (1, 2, 16). In the cardiovascular system, activation of P2 nucleotide receptors exerts potent and various responses including vasodilation, vasoconstriction, and vascular smooth muscle cell proliferation (19). Vascular myocytes express both ionotropic P2X receptors and metabotropic P2Y receptors (19). Four subtypes of P2Y receptors have been found to be coexpressed in the membrane of vascular smooth muscle cells: P2Y₁, P2Y₂, P2Y₄, and P2Y₆ receptors (4). P2Y₂ and P2Y₆ receptors appeared to be the most abundant. P2Y₁ receptor subtype is hardly expressed in contractile vascular smooth muscle cells (4, 26). However, mRNA levels for P2Y₁ receptor, which are undetectable in cells before cultivation, are markedly upregulated in cultured vascular smooth muscle cells (4). All these P2Y receptor subtypes are coupled to phospholipase C activation and rises in intracellular Ca²⁺ concentration ([Ca²⁺]_i) (19, 26). Other signal transduction pathways activated by P2Y receptor stimulation have been proposed, including phospholipase D, phospholipase A2, adenylyl cyclase, and mitogen-activated protein kinase (MAPK) activation (2, 19). Therefore, stimulation of vascular smooth muscle by nucleotides triggers cellular responses that are mediated by more than one P2Y receptor subtype through activation of several intracellular coupling mechanisms. However, the respective roles and coupling mechanisms of these coexpressed P2Y receptor subtypes remain to be understood.

Small GTPases of the Rho family are intracellular signaling proteins known to act as molecular switches to control actin cytoskeleton in fibroblasts (11). Activation of Rho GTPases could be controlled by several mechanisms including the activation of heterotrimeric G protein-coupled receptors. Recently, it was demonstrated that RhoA-dependent signaling pathway can control smooth muscle cell functions such as contraction and proliferation (8, 21, 30). The contracting effect of RhoA results from the activation of Rho-dependent kinase (ROCK), which phosphorylates the myosin light chain phosphatase (15, 20). This leads to the inhibition of its function, thus allowing an increase in the level of phosphorylated myosin light chain and contraction at a constant [Ca²⁺]. This phenomenon is defined as Ca²⁺ sensitization. Coupling between P2Y receptors and the Rho signaling pathway has not been demonstrated. However, it was reported recently that effects of ADP in platelets are sensitive to a Rho kinase inhibitor (27).

In the present study, we thus sought to 1) assess the coupling between P2Y receptors and RhoA-dependent signaling pathway and 2) determine whether P2Y receptor activation triggers Ca²⁺ sensitization in vascular smooth muscle. We present evidence that P2Y receptor stimulation induced activation of RhoA, leading to actin cytoskeleton organization and Ca²⁺ sensitization in a pertussis toxin-independent manner. We identified the P2Y receptor subtypes coupled to RhoA activation as P2Y₁, P2Y₂, P2Y₄, and P2Y₆.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Smooth muscle cell culture. Smooth muscle cells from rat aortas were isolated by enzymatic dissociation as previously described (10). Cells were cultured in DMEM with 10% FCS, 100 U/ml penicillin, and 100 µg/ml streptomycin. Secondary cultures were obtained by serial passages after the cells were harvested with 0.5 g/l trypsin and 0.2 g/l EDTA (Trypsin-EDTA) and then reseeded in fresh DMEM containing 10% FCS and antibiotics.

Measurements of Rho distribution. Serum-starved cultured aortic smooth muscle cells (passages 1 and 2) were washed twice with physiological saline solution [PSS; containing (in mM) 130 NaCl, 5.6 KCl, 1 MgCl₂, 2 CaCl₂, 11 glucose, and 10 Tris; pH 7.4 with HCl] and changed to PSS with or without agonist for indicated periods of time at 37°C. The reaction was stopped by washing the plates with ice-cold PSS, followed by the addition of lysis buffer containing (in mM) 20 HEPES-NaOH, 10 KCl, 10 NaCl, 5 MgCl₂, 1 dithiothreitol and Complete (Boehringer; 1 tablet/50 ml). Plates were placed on ice and then scraped and homogenized. Nuclei and unlysed cells were removed by low-speed centrifugation. The supernatant was then centrifuged at 100,000 g for 30 min to generate membrane and cytosolic fractions. The membrane pellet was resuspended in the same buffer, protein concentration was measured and adjusted, and then Laemmli sample buffer was added. Similar amounts of total proteins were loaded in each lane, electrophoresed on 12% polyacrylamide-SDS gels, and transferred to nitrocellulose. The amounts of proteins were checked by staining with ponceau red. Before immunoblotting was performed, the membrane was blocked with 50 mM Tris · HCl, pH 7.5, 150 mM NaCl, 0.1% Tween 20, and 5% nonfat milk for 1 h at room temperature and then probed with anti-RhoA antibodies (2 µg/ml) for 3 h at room temperature. After three washes, membranes were incubated for 1 h at room temperature with horseradish peroxidase-conjugated goat anti-mouse antibodies (16 ng/ml). The signal from immunoreactive bands was detected by enhanced chemiluminescence and quantified using ImageQuant (Molecular Dynamics).

Actin staining. After dissociation, aortic myocytes were cultured in DMEM with 10% FCS on glass coverslips for 2 days. The cells were then washed and maintained in serum-free DMEM for 48 h. Serum-starved cells were stimulated with P2Y receptor agonist for 40 min in the presence of -MeATP (100 µM, added 1 h earlier), which was used to desensitize any eventual contaminating P2X-induced response. This -MeATP treatment had no effect on actin cytoskeleton organization and Rho activation. Cells were then fixed for 30 min in 4% paraformaldehyde, permeabilized in 0.5% Triton X-100, and rinsed in PBS. For polymerized (F) actin staining, cells were incubated with FITC-conjugated phalloidin (5 µg/ml) for 45 min at room temperature and then washed with PBS. Actin staining has also been performed with a monoclonal anti--smooth muscle actin antibody revealed with FITC-conjugated anti-mouse antibody. Results were similar to those obtained with FITC-conjugated phalloidin. When dual labeling was performed, cells were simultaneously stained with FITC-conjugated phalloidin and Texas Red-labeled DNase I (10 µg/ml) to localize monomeric G-actin (17) and then washed in PBS. Coverslips were mounted on a glass slide and examined with a fluorescence microscope (Eclipse E-600, Nikon). The background fluorescence signal was estimated by collecting planes from areas of the slide without cells and was subtracted before analysis. Images were collected with a cool-SNAP camera (Princeton Instruments) and stored and analyzed using Metamorph software (Universal Imaging, West Chester, PA). For each area examined, images of FITC-phalloidin and Texas Red-DNase I fluorescence were collected. The time of measurements, the time of image capturing, and the image intensity gain at both wavelengths were optimally adjusted and kept constant. The ratio of fluorescence of FITC-phalloidin and Texas Red-DNase I (ratio of F- to G-actin) was calculated for at least 20 cells in each experimental condition. An increase in the ratio of F- to G-actin indicated an increase in stress fiber formation.

Tension measurements in intact fibers. Wistar rats (150 g) were stunned and then killed by cervical dislocation. The aorta was collected in PSS, cleaned of fat and adherent connective tissue, and cut in rings. The endothelium was carefully removed from aortic rings by gently rubbing the intimal surface with the tip of a small forceps. Smooth muscle strips or rings were then suspended under isometric conditions in organ baths filled with Krebs-Henseleit solution (in mM: 118.4 NaCl, 4.7 KCl, 2 CaCl₂, 1.2 MgSO₄, 1.2 KH₂PO₄, 25 NaHCO₃, and 11 glucose), maintained at 37°C, and gassed with 95% O₂-5% CO₂. The preparations were initially placed under a resting tension of 1,500 mg, left to equilibrate for 1 h, and washed at 20-min intervals. The absence of endothelium was confirmed in each ring by the inability of carbachol (10 µM) to relax phenylephrine (PE; 1 µM)-induced contraction. Stimulation with P2Y receptor agonists was performed after desensitization of P2X receptor by 100 µM -MeATP. Concentration-response curves to agonists were obtained by increasing the concentration in the organ chamber, and the maximal tension recorded for each concentration was expressed as a percentage of the maximal response elicited by 1 µM PE.

Expression of P2Y receptors. Full-length P2Y₁, P2Y₂, P2Y₄, and P2Y₆ were cloned in pcDNA3 vector (Invitrogen). Plasmids were transfected into Swiss 3T3 cells using the polyethylenimine (PEI) method as previously described (29). Swiss 3T3 cells were grown on glass coverslips. When cells reached 60-70% confluence, they were washed in FCS-free DMEM. We gently mixed 0.75 µg of P2Y plasmid and 0.25 µg of CD8 plasmid with PEI (0.3 mM) in 50 µl of NaCl (150 mM). The mixture was left to equilibrate for 10 min at room temperature and was then added to the cells bathed in 500 µl of FCS-free DMEM. After 2 days, cells were treated with purinoceptor agonists as described in Actin staining. Anti-CD8 antibody-coated beads were added just before fixation to visualize transfected cells (14). Cells were then fixed and stained for filamentous actin as described in Actin staining.

Overexpression of inactive RhoA mutant. Aortic myocytes grown on coverslips were transiently transfected with a plasmid encoding the inactive mutant V14RhoA(Y35A/T37A) together with the CD8 plasmid using Fugene reagent (Boehringer). Forty-eight hours after transfection, cells were stimulated with purinoceptor agonists. Anti-CD8 antibody-coated beads were added just before fixation to visualize transfected cells, and cells were then fixed and stained for filamentous actin as described in Actin staining.

Statistics. All results are expressed as means ± SE, where n is the sample size. Significance was tested by means of Student's t-test. Probabilities <5% (P < 0.05) were considered significant. Concentration-response curves were fitted to a logistic equation using Origin software (Dipsi, Chatillon, France).

Chemicals and drugs. Texas Red-DNase I was obtained from Molecular Probes (Leiden, The Netherlands). Anti-CD8 antibody-coated beads were purchased from Dynal (Compiègne, France). RhoA antibody (26C4) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The exoenzyme C3 was kindly provided by Dr. P. Boquet (Institut National de la Santé et de la Recherche Médicale U452, Nice University Medical School, Nice, France). The Rho kinase inhibitor Y-27632 was a gift from Yoshitomi Pharmaceutical Industries. All other reagents were purchased from Sigma (Saint Quentin Fallavier, France).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ATP, UTP, and UDP stimulations of aortic smooth muscle cells translocate RhoA protein to the membrane. RhoA activation is known to be associated with an increase in the amount of membrane-bound RhoA. To investigate the ability of P2Y receptor agonists to activate Rho A in aortic smooth muscle cells, we measured changes in membrane-associated RhoA. The -adrenoceptor agonist PE, known to activate RhoA (7), increased the amount of membrane-associated RhoA (1.7-fold increase over the untreated control, Fig. 1) and was used as a positive control. Stimulation of serum-starved aortic smooth muscle cells (passages 1 and 2) with ATP, UTP, or UDP (100 µM) for different periods of time led to an increase in the amount of RhoA in the membrane fraction (Fig. 1). Stimulation with ATP produced a maximal translocation of RhoA to the membrane after 5 min (1.8-fold increase over the untreated control), and then the level of membrane-bound RhoA gradually decreased. In the presence of UTP or UDP, RhoA continued to translocate after 5 min, reaching a peak in the amount of membrane-associated RhoA (2.5-fold increase over the untreated control) after 40 min of stimulation. Stimulation with ADP did not produce a significant increase in the amount of membrane-bound RhoA in myocytes at early passages. However, a twofold increase in membrane-bound RhoA was induced by stimulation with ADP (100 µM, 40 min) in aortic myocytes at passage 5 (not shown).

View larger version (39K): [in this window] [in a new window]   Fig. 1.   Effect of P2Y agonist stimulation on levels of membrane-associated RhoA. Serum-starved aortic myocytes (passages 1 and 2) were stimulated with UTP (100 µM; ), UDP (100 µM; ), or ATP (100 µM; ) for 5-90 min (time 0 corresponds to control conditions in absence of agonists) or with 10 µM phenylephrine (PE; 30 min) used as a positive control. Cells were lysed, scraped, and fractionated as described in text before SDS-PAGE and Western blot analysis with antibody against RhoA. A: representative blots demonstrating time-dependent increase in membrane-associated RhoA in stimulated cells. B: time course of P2Y receptor stimulation-induced increase in membrane-associated RhoA. Amount of RhoA in membrane fraction at time 0 was designated as 1. Data represent means of 3 experiments.

P2Y receptor agonists induced actin cytoskeleton organization in aortic smooth muscle cells. The above results indicate that stimulation of P2Y receptors of aortic smooth muscle cells is coupled to activation of RhoA. We therefore analyzed the ability of the P2Y receptor agonists to induce formation of actin stress fibers in aortic smooth muscle cells. Figure 2 shows serum-starved aortic smooth muscle cells stained with FITC-phalloidin to visualize actin filaments in control and stimulated cells. In untreated cells, only a few thin actin filaments were observed (Fig. 2A). After stimulation with ATP, UDP, or UTP (100 µM; Fig. 2, C, E, or G, respectively), actin filaments appeared as a dense and organized network of thick and parallel fibers. Significant effects were detected at a nucleotide concentration of 1 µM, and maximal effects were obtained at 100 µM. These effective nucleotide concentrations are therefore in the range of the nucleotide concentrations that could be found in the extracellular space after their release from nucleotide-releasing cells (sympathetic neurons, endothelial cells, or platelets) or damaged cells (2). The maximal P2Y receptor agonist-induced stimulation of actin stress fiber formation was quantified and expressed as a 1.9 ± 0.2-fold increase in the ratio of F- to G-actin for ATP (n = 10, P < 0.001), a 3.4 ± 0.2-fold increase for UTP (n = 10, P < 0.001), and a 2.6 ± 0.3-fold increase for UDP (n = 10, P < 0.001) (Fig. 3). The maximal effect of P2Y receptor activation on actin fiber organization has been observed after 40 min of stimulation with extracellular nucleotides. This time was therefore used for all subsequent experiments. The time course of the effect of P2Y receptor stimulation on actin cytoskeleton organization correlates with the time course of RhoA translocation.

View larger version (110K): [in this window] [in a new window]   Fig. 2.   P2Y receptor activation induces actin stress fiber formation in aortic smooth muscle cells. F-actin staining was performed using FITC-phalloidin in unstimulated serum-starved aortic smooth muscles cells (passage 1) (A and B) and in cells stimulated by ATP (100 µM; C and D), UDP (100 µM; E and F), or UTP (100 µM; G and H) in absence (A, C, E, and G) or presence of Rho kinase inhibitor Y-27632 (10 µM; B, D, F, and H). Scale bar represents 10 µm.

View larger version (60K): [in this window] [in a new window]   Fig. 3.   Increase in ratio of F- to G-actin induced by P2Y receptor activation was antagonized by inhibition of RhoA and Rho kinase. A: serum-starved aortic smooth muscle cells (passage 1 and 2) were treated with indicated inhibitors for 1 h, except C3 exoenzyme and pertussis toxin (PTX), which were added to cells for 48 h. Unstimulated cells and cells stimulated with ATP (100 µM), UTP (100 µM), or UDP (100 µM) were stained for both F- and G-actin using FITC-phalloidin and Texas Red-DNase I. Ratio of F- to G-actin fluorescence was then determined in each condition. An increase in ratio of F- to G-actin indicates stress fiber formation. Results are normalized to ratio of F- to G-actin measured in unstimulated control cells. *P < 0.001; ***P < 0.0001. B: phase-contrast photomicrographs of cultured aortic smooth muscle cells (a) showing cell shape change induced by 10 µM Y-27632 (b). Scale bar represents 10 µm.

View larger version (45K): [in this window] [in a new window]   Fig. 4.   Effects of ADP on actin cytoskeleton organization in aortic smooth muscle cells. A: stimulation of actin cytoskeleton organization, quantified by ratio of F- to G-actin, appeared and increased in aortic smooth muscle cells in culture between passages 2 and 5. Results are expressed as a percentage of ratio of F- to G-actin determined in unstimulated serum-starved aortic muscle. F-actin staining was performed using FITC-phalloidin in unstimulated serum-starved aortic smooth muscles cells (passage 3; B) and in cells stimulated by ADP (100 µM; C and D) in absence (B and C) or presence of Rho kinase inhibitor Y-27632 (10 µM; D). Scale bar represents 10 µm. E: serum-starved aortic smooth muscle cells (passages 3 and 4) were treated with indicated inhibitors for 1 h, except C3 exoenzyme and PTX, which were added to cells for 48 h. Unstimulated cells and cells stimulated with ADP (100 µM) were stained for both F- and G-actin using FITC-phalloidin and Texas Red-DNase I. Ratio of F- to G-actin fluorescence was then determined in each condition. *P < 0.001; ***P < 0.0001.

Intracellular signaling coupling P2Y receptor activation to actin cytoskeleton organization. Various pharmacological tools have been used to identify the intracellular signaling pathway coupling P2Y receptor activation to stress fiber formation. Pertussis toxin treatment (5 ng/ml, 48 h) reduced the basal ratio of F- to G-actin by 53.9 ± 8.1% (n = 8, P < 0.001) but did not modify ATP-, UTP-, or UDP-induced actin stress fiber formation (Fig. 3). The involvement of RhoA has been analyzed by treatment of the cells with C3 exoenzyme, which is known to inactivate this small GTPase. After C3 pretreatment (15 µg/ml, 48 h), the cell shape was strongly altered as described for other cell types, and the actin cytoskeleton under basal conditions was almost completely disorganized (Fig. 3). In addition, C3 pretreatment abolished the P2Y agonist-induced actin stress fiber formation (Figs. 3 and 4E). Genistein (20 µM, 1 h), lavendustin (20 µM, 1 h), and herbimycin A (10 µM, 1 h) were used to distinguish whether tyrosine phosphorylation is required for P2Y agonist-induced increase in stress fiber formation in aortic smooth muscle cells. Although these inhibitors reduced the basal ratio of F- to G-actin, none of them affected the P2Y receptor agonist-induced increase in stress fiber formation and the rise in the ratio of F- to G-actin induced by ATP, UTP, or UDP (Fig. 3) or by ADP (Fig. 4E). Similarly, the phosphatidylinositol (PI) 3-kinase inhibitor wortmannin (100 nM, 1 h) decreased the basal ratio of F- to G-actin but had no effect on P2Y receptor activation-induced stimulation of actin stress fiber organization and the rise in the ratio of F- to G-actin (Figs. 3 and 4E). PD-98059 (10 µM, 1 h) and GF-109203 (100 nM, 1 h) were used to test the involvement of MAPKs and protein kinase C, respectively, in the P2Y receptor-induced actin cytoskeleton organization. None of these compounds modified the ATP-, UTP-, UDP-, or ADP-induced stress fiber formation and the rise in the ratio of F- to G-actin (Figs. 3 and 4E). On the contrary, in cells treated with the Rho kinase inhibitor Y-27632 (10 µM, 2 h), the basal F-actin staining (Figs. 2B, 3, and 4E), the P2Y receptor agonist-induced stimulation of stress fiber formation (Fig. 2, D, F, and H and Fig. 4D), and the rise in the ratio of F- to G-actin were abolished (Figs. 3 and 4E). As in C3-treated cells, the shape of the cell was strongly modified, but some short and thin actin fibers persisted. The inhibitory action on actin fibers and the change in the cell shape induced by Y-27632 in cells under control conditions were observed at all passages, indicating that a basal activity of Rho kinase may participate in actin cytoskeleton organization and cultured smooth muscle cell shape. This typical cell shape induced by Y-27632 is shown in Fig. 3B. Taken together, these results suggest that P2Y receptor activation-induced actin cytoskeleton organization in aortic smooth muscle involved RhoA and Rho kinase activation. This coupling mechanism seems to use an intracellular pathway independent of tyrosine kinases, MAPKs, protein kinase C, and PI 3-kinase. However, additional experiments specifically designed to examine each of these intracellular effectors are necessary to confirm this finding. The involvement of RhoA in P2Y receptor activation-induced actin cytoskeleton organization was further analyzed by overexpressing an inactive mutant of RhoA [V14RhoA(Y35A/T37A)] in aortic smooth muscle cells. The inactive mutant of RhoA was coexpressed with CD8, and then transfected cells were identified with the use of anti-CD8 antibody-coated beads. The efficiency of transfection was estimated to 10%. Transfection of CD8 alone did not modify actin cytoskeleton organization in control cells and in cells stimulated with extracellular nucleotides. In all cells expressing the mutant of RhoA, stimulation by extracellular nucleotides did not induce stress fiber formation (Fig. 5), confirming the involvement of RhoA in the P2Y receptor activation-induced responses.

View larger version (140K): [in this window] [in a new window]   Fig. 5.   Overexpression of inactive RhoA mutant in aortic myocytes. Aortic smooth muscle cells were transfected with CD8 and V14RhoA(Y35A/T37A). Transfected cells were identified using anti-CD8 antibody-coated beads. Control cells (A) or cells stimulated with UDP (100 µM; B) or UTP (100 µM; C) were labeled with FITC-phalloidin to reveal polymerized actin (Aa-Ca). Images of same cells focused on beads are shown in Ab-Cb to identify V14RhoA(Y35A/T37A)-expressing cells. P2Y receptor agonist stimulation did not induce stress fiber formation in cells expressing inactive RhoA mutant. Scale bar represents 10 µm.

P2Y receptor agonist-induced Rho kinase-dependent contraction. To further analyze the biological function of the P2Y receptor activation-induced RhoA activation, vasoconstrictor effects of P2Y receptor agonists have been analyzed in endothelium-denuded rings of rat aorta. Application of increasing concentrations of UTP or UDP from 1 µM to 1 mM elicited a dose-dependent rise in tension. UTP and UDP induced maximal contraction of similar amplitudes (Fig. 6). The concentrations producing a half-maximal effect (EC₅₀) were similar for UDP and UTP: 45.1 and 53.7 µM, respectively. It was previously shown that responses to UDP added to cells or tissues may be caused by the conversion of UDP to UTP by the action of extracellular nucleoside diphosphokinase activity as well as the presence of UTP contaminating the supplies of UDP (23). Therefore, the effects of UDP have been examined under conditions in which hexokinase and glucose were included to ensure that no contribution of UTP could occur (23). However, incubation of UDP stock solution with hexokinase and glucose followed by coapplication of UDP with hexokinase and glucose during stimulation of the aortic strips did not modify the UDP-induced contraction, suggesting that UTP contaminant is not responsible for the UDP-induced response (not shown). After desensitization of P2X receptors, application of ATP or ADP in a similar range of concentrations to denuded aortic muscular strips did not produce any rise in tension. The absence of a contracting effect of ATP and ADP in denuded aortic muscle strips could be due to both degradation by ectonucleotidases and subsequent activation of relaxant A_2A receptors and the absence or low level of P2Y₁ receptors in contractile smooth muscle (see DISCUSSION).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Contracting effect of P2Y receptor agonists in aortic smooth muscle is inhibited by Y-27632. A: typical traces show contractile response to increasing concentrations of UTP in deendothelialized rings of rat aorta in absence (left) and presence (right) of Rho kinase inhibitor Y-27632 (1 µM). B: concentration-response curves for contracting effect of UTP in absence () and presence of 1 µM Y-27632 (). C: concentration-response curves for contracting effect of UDP in absence () and presence of 1 µM Y-27632 (). Tension was expressed as a percentage of maximal tension recorded in response to 1 µM PE. Each point represents a mean ± SE of 4-8 experiments.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Y-27632 concentration-dependently inhibited P2Y receptor agonist-induced rise in tension. A: typical traces show inhibitory effect of increasing concentrations of Y-27632 on contraction induced by 300 µM UDP in deendothelialized rings of rat aorta. Concentration-response curves show inhibitory effect of Y-27632 on tension rise induced by 300 µM UTP (B) or 300 µM UDP (C) under control condition () or after treatment of rings with 2 µM thapsigargin and 20 µM D600 (). This treatment prevents Ca2+ release from intracellular Ca2+ store and Ca2+ entry through voltage-gated Ca2+ channel, allowing development of tension response independent of an increase in intracellular [Ca2+]. Tension was expressed as a percentage of maximal amplitude of response recorded before application of Y-27632. Each point represents a mean ± SE of 4-8 experiments.

View larger version (60K): [in this window] [in a new window]   Fig. 8.   Absence of effect of tyrosine kinase, mitogen-activated protein kinase, protein kinase C, and phosphatidylinositol (PI) 3-kinase inhibitors on P2Y receptor agonist-induced rise in tension. Deendothelialized muscle rings of aorta were stimulated by 2 successive applications of UTP (300 µM) or UDP (300 µM) separated by 1 h. The second tension response obtained in absence (control) or presence of indicated inhibitors was expressed as a percentage of the first response. Experiments were performed in presence of 2 µM thapsigargin and 20 µM D600. None of these kinase inhibitors affected contractile response to P2Y receptor agonists. Genistein and lavendustin, tyrosine kinase inhibitors; wortmannin, PI 3-kinase inhibitor; PD-98059, MAPK inhibitor; GF-109203X, protein kinase C inhibitor.

P2Y receptor subtype expression in Swiss 3T3 fibroblasts. The above results indicate that P2Y receptors expressed in aortic smooth muscle cells are coupled to actin cytoskeleton organization through a Rho/Rho kinase-dependent pathway. To identify the subtypes of P2Y receptor responsible for this effect, we expressed recombinant P2Y receptors in Swiss 3T3 fibroblasts and then stained cells to show actin filaments. Each P2Y receptor subtype was coexpressed with CD8. Therefore, transfected cells could be easily identified with the use of anti-CD8 antibody-coated beads. Figure 9 reveals that stimulation with purinoceptor agonists induced stress fiber formation in cells expressing P2Y₁ (Fig. 9, C and D), P2Y₂ (Fig. 9, E and F), P2Y₄ (Fig. 9, G and H), or P2Y₆ (Fig. 9, I and J) receptor subtypes but not in nontransfected cells, indicating that this effect was not mediated by endogenous P2Y receptors. Stimulation of stress fiber also was not observed in cells only transfected with CD8 (Fig. 9, A and B) and in transfected cells in the absence of extracellular nucleotides. Stimulation of actin organization was observed in response to ADP, but not ATP or UDP, in cells expressing P2Y₁ receptor. In cells expressing P2Y₂ or P2Y₄ receptor, stimulation of actin stress fiber was observed in response to UTP and ATP but not UDP. In cells expressing P2Y₆ receptor, stimulation of actin stress fiber formation was induced by UDP but not UTP and ATP. The stimulation of stress fiber formation induced by activation of the recombinant P2Y receptor subtypes in Swiss 3T3 fibroblast was completely antagonized by the Rho kinase inhibitor Y-27632 (10 µM).

View larger version (91K): [in this window] [in a new window]   Fig. 9.   P2Y receptor subtype expression in Swiss-3T3. Swiss-3T3 fibroblasts were transfected with CD8 alone (A and B) or coexpressed with P2Y1 (C and D), P2Y2 (E and F), P2Y4 (G and H), or P2Y6 (I and J) receptor subtypes. Cells were then stimulated with ADP (30 µM; C and D), UTP (30 µM; E-H), or UDP (30 µM; I and J), fixed, and labeled with FITC-phalloidin to reveal polymerized actin. Transfected cells were identified using anti-CD8 antibody-coated beads. Images focused on beads are shown in B, D, F, H, and I. P2Y agonist stimulation induced stress fiber formation in cells expressing P2Y1 (C), P2Y2 (E), P2Y4 (G), or P2Y6 (J) receptor subtypes. Scale bar represents 10 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Our results show that P2Y receptors are coupled to RhoA activation. We also demonstrate with the use of the inhibitor Y-27632 that Rho kinase is the downstream effector responsible for P2Y receptor activation-induced actin cytoskeleton organization and Ca²⁺ sensitization of the contractile apparatus in aortic smooth muscle. This is the first demonstration that RhoA and Ca²⁺ sensitization are activated by extracellular nucleotides. Considering the existence of multiple downstream target molecules and the pleiotropic function of Rho including cytoskeleton rearrangements, stimulation of DNA synthesis, cell proliferation, and migration (22), RhoA appears as a new key element in the intracellular signaling pathways responsible for the numerous effects of extracellular nucleotides in vascular smooth muscle.

Much experimental evidence supports the activation of RhoA by P2Y receptor stimulation. First, P2Y receptor agonists cause translocation of a fraction of Rho from the cytosolic to the particulate fraction. Quantitatively, this RhoA translocation is similar to that reported for other agonists of G protein-coupled receptors such as norepinephrine (7) and thrombin (30). Second, purinoceptor agonist-induced actin stress fiber formation was inhibited by C3 treatment. Third, overexpression of inactive mutant of RhoA (Y35A/T37A) in aortic smooth muscle cells prevented P2Y receptor activation-induced actin cytoskeleton organization. Taken together, all these results are consistent with RhoA activation by extracellular nucleotides in vascular smooth muscle.

Actin cytoskeleton organization can be triggered by ATP, ADP, UTP, and UDP in a Rho kinase-dependent manner, suggesting that potentially all the mammalian P2Y receptor subtypes known to be expressed in cultured rat aortic smooth muscle cells (P2Y₁, P2Y₂, P2Y₄, and P2Y₆) could be coupled to the RhoA-dependent signaling pathway. Expression of these receptors in Swiss 3T3 cells confirmed this hypothesis because actin stress fiber formation was induced by extracellular nucleotides in cells expressing P2Y₁, P2Y₂, P2Y₄, or P2Y₆ receptor subtype through a Rho kinase-dependent pathway. In addition to RhoA translocation and actin cytoskeleton organization, UTP and UDP also induced a Rho kinase-dependent contraction in rings of rat aorta. The UTP- or UDP-induced contractile response that was independent of [Ca²⁺]_i rise recorded in the presence of D600 and TSG was completely inhibited by the Rho kinase inhibitor Y-27632, suggesting that UTP and UDP induced a Rho/Rho kinase-dependent Ca²⁺ sensitization. On the contrary, although ATP induced RhoA translocation and actin cytoskeleton stimulation in cultured aortic smooth muscle cells, we were not able to measure a significant contractile response to ATP in aortic muscle rings. This absence of a contracting effect of ATP could result from degradation by ectonucleotidases (28, 34), leading to production of adenosine and subsequent activation of relaxant A_2A receptors in smooth muscle cells, which compensated the contracting effect of ATP (9). Similarly, although ADP induced Rho kinase-dependent actin cytoskeleton organization in cultured aortic smooth muscle cells, it did not contract aortic smooth muscle. This observation agreed with the absence of rise in [Ca²⁺]_i in response to ADP or other P2Y₁ receptor agonists in freshly isolated aortic smooth muscle cells (26) and the absence of actin cytoskeleton stimulation in cultured myocytes at early passages. Together with the absence of mRNA for P2Y₁ receptor in contractile smooth muscle (4), these results confirm that P2Y₁ receptor was hardly expressed or not expressed at all in aortic smooth muscle cells and suggest that the P2Y₁ receptor subtype was not involved in the activation of RhoA triggered by extracellular nucleotides in aortic smooth muscle. In aortic smooth muscle rings, Rho kinase-dependent Ca²⁺ sensitization was induced in response to UTP or UDP even in the presence of hexokinase, suggesting that P2Y₂, P2Y₄, and P2Y₆ receptor subtypes can all participate in the stimulation of RhoA- and Rho-dependent intracellular pathways in vascular smooth muscle.

The coupling mechanisms between heterotrimeric G protein-coupled receptors and Rho activation are not yet totally understood. It was shown recently (6, 13) that C3-sensitive actin cytoskeleton organization can involve G_i, G_q/G₁₁, G₁₂, or G₁₃. In aortic smooth muscle, the Rho-dependent stimulation of actin stress fiber formation was not sensitive to pertussis toxin treatment, suggesting that the heterotrimeric G protein(s) involved did not belong to the G_i family. Although P2Y receptors are known to be coupled to G_q proteins, G₁₂ and G₁₃ also appeared to be good candidates for coupling P2Y receptors to Rho activation, because guanine nucleotide exchange factors for Rho selectively and directly activated by G₁₂ and G₁₃ were recently identified (5, 12, 18). In addition, it was shown (24) in fibroblasts that agonists such as lysophosphatidic acid activate RhoA through a tyrosine kinase-dependent pathway (24). In aortic smooth muscle cells, although tyrosine kinase inhibitors decreased the F-actin organization under unstimulated conditions, they did not prevent the Rho kinase-dependent contraction and actin fiber formation induced by extracellular nucleotides, suggesting that tyrosine kinases are not likely to be involved in the activation of the Rho/Rho kinase signaling pathway. The observed effect of tyrosine kinase inhibitors indicates, however, that a tyrosine kinase-dependent pathway also regulates actin cytoskeleton organization in vascular smooth muscle, independently of soluble extracellular agonists.

Many studies have reported the multiple effects of extracellular nucleotides in the cardiovascular system (19). However, the intracellular pathways responsible for the biological responses induced by extracellular nucleotides have not been completely identified. Nucleotides have been shown to modulate vasomotricity but could also be involved in pathophysiological processes such as atherosclerosis, neointima formation after angioplasty, and hypertension (3). The demonstration of the coupling between P2Y receptors and RhoA activation gives a new opening to elucidate the signaling mechanisms involved in the P2Y receptor activation-induced cellular responses. Indeed, numerous effects of extracellular nucleotides such as cell contraction, proliferation, and migration are also known to be induced by RhoA activation (22, 33). Further studies are now required to identify downstream effectors of Rho other than Rho kinase that are activated by P2Y receptor stimulation and to define among the nucleotide-induced cellular responses those depending on RhoA activation. Nevertheless, inhibition of RhoA or Rho kinase appears to be a new pharmacological strategy to prevent or limit smooth muscle cell proliferation and vasoconstriction associated with arterial diseases such as atherosclerosis and hypertension. Inhibition of Rho kinase is expected to reduce phenomena linked to cell morphology, contraction, or migration, and the Rho kinase inhibitor Y-27632 has been shown to restore normal arterial pressure in animal models of hypertension (32). RhoA inhibition should, in addition, decrease smooth muscle cell proliferation. The inhibition of vascular smooth muscle cell growth induced by statins (31), which prevent isoprene synthesis and therefore prenylation of Rho proteins, is in agreement with this proposal and suggests that specific of RhoA inhibitors deserve to be found.